Google Research used its Research@ 2026 event to present AI systems aimed at extending human capabilities in science, health, and hardware. Among the highlights is Google Earth AI, which combines planetary foundation models with tools like Dataset Search and TimesFM to support geospatial analysis, risk assessment, and custom workflows. Google.org is also committing $10 million to organizations working on crisis resilience and public health.
For scientific work, Google described a multi-agent orchestrator that lets researchers reproduce prior methods and adapt them to new challenges. One example applies a fluid-mechanics methodology to urinary catheter design, finding simulated designs that better resist bacterial infections. Separately, Deep Review, an agentic system already used on more than 16,000 research submissions, is now available to the public for checking technical documents such as legal agreements and financial reports.
On the hardware side, Google highlighted a co-designed next-generation low-power Edge AI chip built by Synaptics using Google's open source IP. The chip is intended to help engineers build smart glasses, watches, and connected cameras with all-day performance. The source is a single Google Research announcement, so no independent or contrasting perspectives are included.